Safety and Security Engineering

405

Increasing Efficiency in Airport Security

Screening
A. Schwaninger
Department of Psychology, University of Zurich, Switzerland

Abstract
The prominence of aviation security has increased dramatically in recent years.
As a reaction to the new threat situation large investments into modern security
technology have been made. State-of-the-art x-ray screening equipment provides
high resolution images and many image enhancement features. However, the
most expensive equipment is of limited value if the humans who operate it are
not selected and trained to perform their task accurately and efficiently. This
article presents results of studies conducted over the last four years which clearly
show that efficiency in x-ray screening can be increased substantially by
investing into human factors technologies that are based on results of visual
cognition, object recognition and psychophysics.
Keywords: Airport security, human factors, detection performance, computerbased training, pre-employment assessment

Introduction

X-ray screening of passenger bags is an essential component of aviation security.

Baggage screening technologies have evolved substantially over the last decades
and state-of-the art x-ray systems can provide high quality images in a good
resolution. However, recognizing object shapes in x-ray images is still a very
difficult problem for computers. Any serious computer vision scientist will tell
you that there still is a long way to go in order to build an artificial visual
recognition system that can detect objects as quickly and reliably as humans do
(for recent advances in biologically motivated computer vision see for example
Blthoff, Lee, Poggio, & Wallraven [1]). Detecting all kinds of prohibited items
quickly and reliably in x-ray images will continue to depend critically on human
operators. Since there are large differences between people in terms of their
WIT Transactions on The Built Environment, Vol 82, 2005 WIT Press
www.witpress.com, ISSN 1743-3509 (on-line)

406

Safety and Security Engineering

visual abilities, a reliable object recognition test could be a critical component of

a pre-employment assessment system (see next section). Moreover, screeners
must know which objects are prohibited and what they look like in passenger
bags. As will be shown in section 3, individually adaptive computer-based
training can be a very effective and efficient tool in order to achieve and
maintain a highly reliable and fast detection of threat items by human operators.

Image-based factors and screener ability

Several image-based effects influence how well threat items can be recognized in
x-ray images (Schwaninger [2]; Schwaninger, Hardmeier, & Hofer [3]). As
illustrated in Fig. 1a, viewpoint can strongly affect recognition performance.
Systematic effects of viewpoint have been reported in many object recognition
studies (for recent reviews see Graf, Schwaninger, Wallraven & Blthoff [4];
Schwaninger [5]). In general, objects are difficult to recognize when depicted
from an unusual viewpoint and when diagnostic features are not visible. Object
recognition has also been shown to be dependent on exposure. When an object is
not similar to a stored view, it often becomes difficult to recognize.
a Viewpoint

b Superposition

c Bag
Complexity

Figure 1: Image-based factors: a) effect of viewpoint (canonical vs. noncanonical), b) superpositition by other objects (low vs. high), c) bag
complexity (low vs. high).
A second image-based factor is related to the fact that in x-ray images objects are
often superimposed by other objects. As depicted in Fig. 1b this effect of
superposition can impair detection performance substantially. A third factor is
bag complexity, which is determined by the number and type of other objects in
the bag (Fig. 1c). Since other objects distract attention, bag complexity can also
affect detection performance.
In order to measure how well people can cope with these image-based factors,
the x-ray Object Recognition Test (ORT) has been developed (Schwaninger [2]).
WIT Transactions on The Built Environment, Vol 82, 2005 WIT Press
www.witpress.com, ISSN 1743-3509 (on-line)

Safety and Security Engineering

407

This test uses only guns and knives, object shapes that are known well by
novices. It contains 256 x-ray images, half of them with a threat item and half of
them with harmless bags. The guns and knives are depicted in easy and difficult
viewpoint, with little and much superposition in bags of low and high
complexity. All x-ray images of the ORT are in black-and-white, as color is
mainly diagnostic for the material of objects in the bag and thus is not very
helpful for novices. The ORT is fully computer-based. Each image is only
displayed for 4 seconds, simulating conditions of high passenger flow during
which average inspection times often are in the range of 3-5 seconds. For each xray image, participants have to decide whether the bag contains a threat item or
whether it is harmless.
Fig. 2 shows means and standard deviations from a study conducted with 284
candidates that applied for an x-ray screening job. Detection performance is
indicated by A, which can be calculated from hit and false alarm rates (for
details see Schwaninger, Hardmeier, & Hofer [3]; Hofer & Schwaninger [6]).
1.00

Detection Performance (A')

*
0.90

*
*

0.80

0.70

0.60

0.50

Unrotated
View

Rotated
View

Superposition Superposition
Low
High

Complexity
Low

Complexity
High

Figure 2: Effects of viewpoint, superposition, and bag complexity found in a

study conducted with 284 novices that applied for an airport security
screening job. Thin error bars represent standard deviations. Stars
indicate the scores of the 2.5% best and 2.5% worst performing
candidates.
There were systematic effects of view, superposition, and bag complexity. When
the threat items were depicted in rotated view, recognition performance dropped
substantially. When the guns and knives were superimposed by other objects,
detection performance was also reduced significantly. Finally, showing threat
items in complex bags, made them more difficult to recognize.
Really remarkable is the fact that there are very large differences between
individuals with regard to how well they can cope with these image-based
effects. The thin error bars in Fig. 2 depict standard deviations and the stars
indicate the performance of the 2.5% best and 2.5% worst performing
candidates. Similar main effects and large inter-individual differences have been
consistently found for different participant groups involving novices and aviation
WIT Transactions on The Built Environment, Vol 82, 2005 WIT Press
www.witpress.com, ISSN 1743-3509 (on-line)

408

Safety and Security Engineering

security screeners at different European airports. Screeners usually perform

better, but we always found significant main effects of view, superposition and
bag complexity. Most importantly, even for screeners, large differences can be
found between people with regard to how well they can cope with effects of
view, superposition and bag complexity. The results of a recent study conducted
with 134 novices and 134 aviation security screeners suggest that these imagebased effects are related to visual abilities. Schwaninger, Hardmeier, & Hofer [3]
found that people differ substantially with regard to screener ability and this
seems to remain relatively stable even after years of working as a screener. A
recent analysis revealed that the results of the ORT are correlated with detection
performance measured with CTI threat image projection on the job (correlations
of r > .5). This further implies that a substantial increase in detection
performance can be achieved if tests like the ORT are used as part of a preemployment assessment system (which is the case at Zurich airport since 2004).

Computer-based training and screener competency

The previous section has shown that image-based factors such as view,
superposition and bag complexity affect how well threat items can be detected.
Even more important is the knowledge about which items are prohibited and
what they look like in x-ray images of passenger bags. Consider the three x-ray
images depicted in Fig. 3.

Figure 3: Threat detection in x-ray images is strongly dependent on training and

expertise. a) two threat items that are not well known from everyday
experience (self-defense gas spray and torch light with shooting
mechanism), b) threat item that looks very different in the X-ray image
(taser) vs. threat item that looks like an innocent object (switchblade
knife looking like a pen), c) two threat items that are difficult to
recognize without training because they are depicted in unusual views
(gun and scissors).
The bag on the left of Fig. 3 contains two threat items that are difficult to
recognize for novices because most people are not exposed to them in reality.
The bag in the middle contains a threat item that is quite difficult to identify
without training because the x-ray image looks very different than the real
object. In contrast, the other threat item in Fig. 3b is difficult to recognize
WIT Transactions on The Built Environment, Vol 82, 2005 WIT Press
www.witpress.com, ISSN 1743-3509 (on-line)

Safety and Security Engineering

409

because it looks similar to a pen, a well known everyday object. The bag on the
right illustrates another problem that was already mentioned in the previous
section. When objects are depicted in an unusual view, they become difficult to
recognize if view-invariant diagnostic features are not available (for recent
reviews see Graf, Schwaninger, Wallraven & Blthoff [4]; Schwaninger [5]).
Since this is often the case in x-ray images, a large image library in which
objects are depicted in different views is necessary in order to achieve reliable
threat detection despite changes in viewpoint.
3.1 Multiple views TIP library
Essentially, object recognition means to match visual information from stimuli
(e.g. an x-ray image) to visual representations stored in memory. The content of
visual memory is dependent on exposure. Shapes are difficult to recognize if
they are not similar to objects that have been encountered earlier and were stored
in visual memory. This was illustrated in Fig. 3 at the beginning of section 3.

Figure 4: Illustration of different knowledge-based factors using different types

of threat items. a) Forbidden object with an unusual shape (selfdefense gas spray), b) prohibited item that looks very different in the
x-ray image (taser), c) threat object that looks like a harmless object
(switchblade knife looking like a pen), d) and e) effects of viewpoint
illustrated with a gun and scissors.
The six threat items in Fig. 3 are very difficult to recognize due to different
reasons (see also Fig. 4). First, many threat objects are not known from everyday
experience. Examples are the self-defense gas spray and the torch light with
shooting mechanism that are contained in the x-ray image depicted in Fig. 3a and
Fig. 4a. Second, certain threat items look quite different than the real object. An
example is the taser weapon shown in Fig. 3b and Fig. 4b. Moreover, certain
other threat items look very similar to harmless objects. For example the

WIT Transactions on The Built Environment, Vol 82, 2005 WIT Press
www.witpress.com, ISSN 1743-3509 (on-line)

410

Safety and Security Engineering

switchblade knife in Fig. 3b could be easily mistaken with a pen (see also Fig.
4c). A third very important but often neglected problem are effects of rotation.
As mentioned earlier, when objects are depicted in unusual viewpoints, they can
become very difficult to recognize. This was illustrated for a gun and scissors in
Fig. 3c. The upper row in Fig. 4d and Fig. 4e depicts these two threat items in
canonical view, i.e. the view that is easiest to recognize. As you can see in the
bottom row, effects of viewpoint can indeed be detrimental to recognition
performance.
D
i
e
n
s
t
w
a
f
f
e
S
I
G
2
2
5

ho00

ve45

ho45

ve85

ho85

vh45

a Six Basic Views

(fully standardized)

b All views flipped

horizontal and vertical
=24 views

c Each view in
3 planar rotations

=72 Views / Object

Figure 5: Basic principles of the multiple views x-ray image library.

Based on a close collaboration between vision scientists and aviation security
experts we have built a multiple views library that currently contains 36000 xray images of fictional threat items (FTIs). Object recognition research indicates
that for most objects six views are sufficient to capture the qualitative differences
resulting from viewpoint changes. These six views constitute the basic view set
(Fig. 5a). The easiest view is referred to as the canonical view. It can be
determined objectively by analyzing detection performance data from CBT or
TIP. Alternatively, subjective rating methods can also provide a good estimate of
viewpoint difficulty, at least on an ordinal scale. The remaining views of the
basic view set can be created by rotating the object around the vertical axis and
the horizontal axis. The method depicted in Fig. 5a uses 45 and 85 rotations
around either the horizontal or the vertical axis and 45 rotations around both the
horizontal and the vertical axis. Using computer algorithms mirror-reversed
versions can be created as depicted in Fig. 5b. Since rotations in the plane can
also have an effect on recognition performance it is advisable to create plane
rotated versions (Fig. 5c).
The current version of our view-based library contains 500 prohibited items and
each of them can be displayed in 6 basic views x 4 mirror reversals x 3 plane
rotations = 72 views. This results in a total of 500 prohibited items x 72 views =
36000 x-ray images. Of course it is not necessary to show each screener all
these images. But an individually adaptive training system should determine for
each screener which views are difficult and adapt the training so that the trainee
WIT Transactions on The Built Environment, Vol 82, 2005 WIT Press
www.witpress.com, ISSN 1743-3509 (on-line)

Safety and Security Engineering

411

becomes able to detect threat items reliably even if they are substantially rotated
away from the easiest view.
3.2 Individually adaptive training
After analysing the results of the first studies conducted at Zurich Airport in
2000 and 2001 we were looking for a CBT system that has a large image library
containing thousands of images depicting threat items in different viewpoints.
Moreover, such a system should be efficient and effective by applying
individually adaptive training algorithms that calculate for each screener which
items and views need to be trained particularly. The results of our first survey
were that all existing CBT systems had only small image libraries and they did
not feature individually adaptive training algorithms. Instead, the commercially
available systems focussed on providing many image processing features such as
negative image, black and white, edge detection etc. It actually sometimes
seemed that the quality of a CBT would be dependent on providing an exact
replica of the x-ray machine with as many image processing and enhancement
features as possible. However, the goal of the training system we had in mind
was to train screeners to detect threat items reliably within the first 3-5 seconds
of inspection time before image processing features might be applied. In order to
achieve this, a scientifically based training system was needed that features a
large image library, sophisticated adaptive training algorithms and feedback
mechanisms that are based on results of perceptual learning. Since nothing like
that existed, we decided to create such a system on our own. In a close
collaboration between psychologists, vision and computer scientists as well as
aviation security experts the first version of X-Ray Tutor was built in 2002 and
installed at Zurich airport. It is now used by some of the largest security
organizations in the world. The United States Transportation Security
Administration is using X-Ray Tutor at airports across the country since spring
2004 as part of its recurrent training and professional development program for
airport security screeners. The Canadian Air Transport Security Authority has
started using X-Ray Tutor at several airports as part of a research project with the
University of Zurich on x-ray image tutoring technology for aviation security
screeners. X-Ray Tutor is operational in Switzerland since 2002, at 17 airports in
Germany since 2003, and since 2004 in Belgium and other European countries.
At the heart of X-Ray Tutor are results from scientific studies on how the human
brain processes visual information in order to recognize objects from different
viewpoints (Schwaninger [7], [8]; for recent overviews on object recognition see
Graf, Schwaninger, Wallraven & Blthoff [4]; Schwaninger [5]). The system
uses a sophisticated algorithm to create training sessions for each individual
student based on detection performance, response time and x-ray image
difficulty ratings. The individually adaptive algorithm of X-Ray Tutor starts with
easy x-ray images. Then, view difficulty is increased and trainees learn to detect
threat items in different viewpoints. X-Ray Tutor uses the view-based image
library described above, which allows presenting each threat item in 72 different
viewpoints. As mentioned before, it is not necessary to show each item in all
viewpoints. The adaptive algorithm of X-Ray Tutor shows for each trainee
WIT Transactions on The Built Environment, Vol 82, 2005 WIT Press
www.witpress.com, ISSN 1743-3509 (on-line)

412

Safety and Security Engineering

preferentially those views, which need to be trained especially. Once a threat

item can be recognized despite changes of viewpoint, it is shown in more
complex bags with more superposition. The individually adaptive algorithm uses
objective measures of view difficulty, superposition and bag complexity. These
measures have been shown to be correlated with human detection performance
and they are based on several years of research involving vision scientists, image
processing specialists, statisticians, and aviation security experts. As students
progress the algorithm monitors students performance and custom designs each
subsequent training session to target areas in need of improvement.
Students learn more when given the opportunity to receive immediate feedback
about their performance. X-Ray Tutor provides detailed feedback during a
training session about the students performance. Students can review any x-ray
image to learn where the threat item is located and what it looks like both in xray and standard photographic view (see Fig. 6). At the end of each training
session, students are given feedback on their performance, so they can visually
see their improvement over time.

Figure 6: Screenshot of X-Ray Tutor CBS. Immediate feedback is provided after

each trial and trainees can display a threat information window
showing the threat item as x-ray image and as photograph.
X-Ray Tutor can also be used for several research purposes. Using the data of
training trials is complicated by the fact, that the training system adapts to each
individual and displays more difficult images the better a screener becomes in
detecting threats. In order to achieve a more reliable measurement of detection
performance across time, a certain number of test images can be shown within
training sessions. These images are the same for all screeners providing a more
reliable basis for standardized detection performance measurement. The test
images can be shown periodically at criterion-based or time-based intervals so
that an evaluation of training effectiveness across time becomes possible. When
X-Ray Tutor is used in a research project, the data from testing trials is stored in
WIT Transactions on The Built Environment, Vol 82, 2005 WIT Press
www.witpress.com, ISSN 1743-3509 (on-line)

Safety and Security Engineering

413

reports than can be imported easily into applications for further analysis such as
standard office software or analysis tools for large databases. Using the reporting
feature, detection performance can be analyzed and visualized across time, for
individual screeners, groups of screeners, specific FTIs and groups of FTIs. This
data can then be used in order to answer several important questions such as for
example:
- How does detection performance increase over time?
- Which threat items are difficult to recognize?
- How does viewpoint affect threat detection performance?
- How reliable are individual performance measurements?
These questions are currently being investigated in research projects conducted
in our lab. The following section summarizes the results of a study that was
conducted in 2002/2003 in order to measure training effectiveness and
investigate perceptual learning mechanisms. A detailed description of this study
can be found in Schwaninger and Hofer [9].
3.3 Measuring Effectiveness and investigating perceptual learning
The study was conducted in the period of December 2002 May 2003 with 72
participants. None of them had received computer-based training before. Each
week, 1-2 training sessions of 20 minutes were completed during the six months
period. X-Ray Tutor HBS for training detection of improvised explosive devices
(IEDs) was used. The research questions were as follows:
- Does IED detection performance increase over time as a result of individually
adaptive training?
- Are there large differences in detection performance when images are displayed
4 vs. 8 seconds?
- Does training result in a decrease of response time? If so, is that decrease
different for threat images vs. harmless bag images?
Four tests were conducted in which new IEDs were used that had not been
shown previously during training. Display durations at test were 4 and 8 seconds
(during training, images were displayed for a maximum of 12 seconds). The four
tests were of equal difficulty as determined in a pilot study. Signal detection
theory was used to analyze the data (Green & Swets [10]). Detection
performance was measured using d = z(H) z(FA), whereas H represents the hit
rate, FA the false alarm rate and z the z-transformation. Similar results were
obtained using non-parametric A instead of d as performance measure (for a
comprehensive introduction on signal detection and x-ray screening see
Schwaninger [11]; for more detailed information on different detection measures
of x-ray detection performance see Hofer and Schwaninger [6]). The results of
the four tests are displayed in Fig 7. There was only a small effect of display
duration, i.e. detection performance was similar for display durations of 4 and 8
seconds. This is consistent with the view that recognition is a fast process
occurring within the first few seconds of image inspection. More important is the
knowledge about threat items, which is dependent on training. Indeed, large
WIT Transactions on The Built Environment, Vol 82, 2005 WIT Press
www.witpress.com, ISSN 1743-3509 (on-line)

414

Safety and Security Engineering

increases of detection performance were observed across the four months of

training (Fig. 7a). In order to assess training effectiveness we calculated %
increase as compared to baseline measurement (first test results). As you can see
in Fig. 7b relative detection performance was increased by about 71%. This is a
remarkable effect if it is taken into account that on average screeners took only
28 training sessions during the six months period (SD = 10 training sessions).
Moreover, for a subgroup of 52 screeners, who on average took 31 training
sessions (SD = 8 training sessions), the training effect was even more
pronounced. For this group relative detection performance was increased by
84%!
100

4 sec
8 sec

% Increase in Detection d'

Detection Performance d'

80
60
40

4 sec
8 sec

20
0

1
Dec/Jan

Feb/Mar

Dec/Jan

Mar/April April/May

Feb/Mar

Mar/April April/May

Figure 7: a) Absolute detection performance measured by d, b) detection

performance increase expressed in percent relative to the first
measurement. Error bars represent SEM.
The analysis of response times revealed interesting findings with regard to
efficiency. As can be seen in Fig. 8, training with X-Ray Tutor resulted in a
much faster detection of threat items. Correct response times for threat images
dropped from about 5 seconds before training to about 3.5 after six months of
training.
Threat Images

Non-Threat Images
4 sec
8 sec

6
5
4
3
2

Response Time (sec)

Response Time (sec)

4 sec
8 sec

6
5
4
3
2

Dec/Jan

Feb/Mar

Mar/April

April/May

Dec/Jan

Feb/Mar

Mar/April

April/May

Figure 8: Response time data as a function of training. a) Response times for

correct decisions on threat images, b) response times for correct
decisions on non-threat images. Error bars represent SEM.
WIT Transactions on The Built Environment, Vol 82, 2005 WIT Press
www.witpress.com, ISSN 1743-3509 (on-line)

Safety and Security Engineering

415

There was also an effect of display duration, although much smaller in effect size
than the overall training effect. For harmless bags average response times
remained constant at about 5 seconds. This finding is consistent with the
assumption that individually adaptive CBT increases the number and strength of
representations in visual memory. Because the participants have learnt what
IEDs look like they became able to detect them reliably within a few seconds of
inspection time. However, the acquired knowledge does not help if there is no
threat item in the bag. Since no substantial reduction in response time was
observed for harmless bag images, the learning effect seems rather related to
visual memory representations than to increased general visual processing
capacities.

Conclusion

These results show that pre-employment assessment and individually adaptive

training can be very powerful tools in order to increase efficiency in aviation
security screening. Moreover, the last study also provides a nice example for the
value of X-Ray Tutor as a research tool. For measuring detection performance
and training effectiveness, as well as gaining a better understanding of perceptual
learning mechanisms and object recognition.

Acknowledgements

I am thankful to airport and police authorities who provided data and funding for
conducting the studies reported in this paper. Thanks to Franziska Hofer and
Judith Riegelnig for their help in preparing the manuscript.

References

[1] Blthoff, H.H., Lee, S.-W., Poggio, A., & Wallraven, C., Biologically
motivated computer vision, Lecture Notes in Computer Science, 2525,
Berlin: Springer, 2002.
[2] Schwaninger, A., Evaluation and selection of airport security screeners.
AIRPORT, 02/2003, 14-15, 2003.
[3] Schwaninger, A., Hardmeier, D., & Hofer, F., Measuring visual abilities and
visual knowledge of aviation security screeners. IEEE ICCST Proceedings,
38, 258-264, 2004.
[4] Graf, M., Schwaninger, A., Wallraven, C., & Blthoff, H.H.,
Psychophysical results from experiments on recognition & categorisation.
Information Society Technologies (IST) programme, Cognitive Vision
Systems - CogVis (IST-2000-29375), 2002.
[5] Schwaninger, A., Object recognition and signal detection. In: B. Kersten &
M.T. Groner (Eds.), Praxisfelder der Wahrnehmungspsychologie. Bern:
Huber, 2004.

WIT Transactions on The Built Environment, Vol 82, 2005 WIT Press
www.witpress.com, ISSN 1743-3509 (on-line)

416

Safety and Security Engineering

[6] Hofer, F., & Schwaninger, A., Reliable and valid measures of threat
detection performance in x-ray screening, IEEE ICCST Proceedings, 38,
303-308, 2004.
[7] Schwaninger, A., Training of airport security screeners. AIRPORT, 05/2003,
11-13, 2003.
[8] Schwaninger, A., Computer based training: a powerful tool to the
enhancement of human factors. Aviation Security International, FEB/2004,
31-36, 2004.
[9] Schwaninger, A. & Hofer, F. Evaluation of CBT for increasing threat
detection performance in x-ray screening. In: K. Morgan and M. J. Spector,
The Internet Society 2004, Advances in Learning, Commerce and Security
(pp. 147-156). Wessex: WIT Press, 2004.
[10] Green, D. M. & Swets, J. A., Signal detection theory and psychophysics,
Wiley: New York, 1996.
[11] Schwaninger, A., Reliable measurements of threat detection. AIRPORT,
01/2003, 22-23, 2003.

WIT Transactions on The Built Environment, Vol 82, 2005 WIT Press
www.witpress.com, ISSN 1743-3509 (on-line)